Response Surface Methodology to Characterize the Indicators for the Early Detection of Fraudulent Cases in the Motor Insurance Market

The motor liability insurance market currently has many stakeholders, so investigating potential fraud in this area takes top priority. The payments related to fraudulent claims cause, in particular, economic damage, so several solutions have been developed to detect them at an early stage. In this paper, a response surface methodology-inspired concept is introduced to estimate the theoretical assumed probability of being an insurance fraud. For that purpose, three easily and unambiguously identifiable factors have been chosen: the insurance payout, EUR; the age of the innocent participant vehicle, years; and the payment period of the insurance contract. The methodology was applied to select the parameters with the most substantial effect on the fraud risk.